Post on 23-Apr-2021
transcript
Surface Integrity and Fatigue Behaviour of Electric Discharged Machined and Milled
Austenitic Stainless Steel
Mattias Lundberg, Jonas Saarimäki, Johan Moverare and Mattias Calmunger
Journal Article
N.B.: When citing this work, cite the original article.
Original Publication:
Mattias Lundberg, Jonas Saarimäki, Johan Moverare and Mattias Calmunger, Surface Integrity and Fatigue Behaviour of Electric Discharged Machined and Milled Austenitic Stainless Steel, Materials Characterization, 2017. 124, pp.215-222. http://dx.doi.org/10.1016/j.matchar.2017.01.003 Copyright: Elsevier
http://www.elsevier.com/
Postprint available at: Linköping University Electronic Press
http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-134043
1
Surface Integrity and Fatigue Behaviour of Electric Discharged Machined and
Milled Austenitic Stainless Steel
Mattias Lundberga, *, Jonas Saarimäkia, Johan J. Moverarea, Mattias Calmungera
aDivision of Engineering Materials, Department of Management and Engineering, Linköping
University, SE-581 83 Linköping, Sweden
*Corresponding author: mattias.lundberg@liu.se
Abstract
Machining of austenitic stainless steels can result in different surface integrities and different
machining process parameters will have a great impact on the component fatigue life. Understanding
how machining processes affect the cyclic behaviour and microstructure are of outmost importance in
order to improve existing and new life estimation models. Milling and electrical discharge machining
(EDM) have been used to manufacture rectangular four-point bend fatigue test samples; subjected to
high cycle fatigue. Before fatigue testing, surface integrity characterisation of the two surface
conditions was conducted using scanning electron microscopy, surface roughness, residual stress
profiles, and hardness profiles. Differences in cyclic behaviour were observed between the two surface
conditions by the fatigue testing. The milled samples exhibited a fatigue limit. EDM samples did not
show the same behaviour due to ratchetting. Recrystallized nano sized grains were identified at
the severely plastically deformed surface of the milled samples. Large amounts of bent mechanical
twins were observed ~ 5 μm below the surface. Grain shearing and subsequent grain rotation from
milling bent the mechanical twins. EDM samples showed much less plastic deformation at the surface.
Surface tensile residual stresses of ~ 500 MPa and ~ 200 MPa for the milled and EDM samples
respectively were measured.
2
Keywords: Austenitic stainless steel, Fatigue, Surface Integrity, SEM, XRD, Hardness
1. Introduction
Stainless steels are used in various industries such as aerospace, automotive, biomedical, and power
generation. Due to its wide field of use, extensive research has been conducted in order to improve
the mechanical properties and performance of stainless steels. Manufacturing a component to the
required geometrical tolerances without the use of any machining is next too impossible. All types of
machining e.g., milling, turning, drilling, plasma machining, laser cutting, water jet cutting, and
electrical discharge machining (EDM) results in different surface integrities, resulting in different
mechanical behaviour. Optimising machining parameters in order to minimise machining time without
compromising surface quality and/or given geometrical tolerances have been investigated in [1].
Different surface conditions can alter the mechanical behaviour during cyclic loading and influence
fatigue life of AISI 304 and AISI 316 stainless steels [2–13]. To counteract the possibly detrimental
effects that might be induced by machining, post treatments can be used to increase fatigue resistance
and component life. A common way to achieve this is to altering the surface integrity by e.g.,
introducing an increased strain hardened layer, grain fragmentation and/or residual stress (RS)
optimisation [7,14–16]. Commonly used surface processing techniques in industry today are shot
peening, laser shock peening, surface mechanical attrition treatment and deep rolling. Machining and
surface treatments can alter mechanical properties such as fatigue strength. Therefore, it is essential
to understand the microstructural evolution at the surface and its impact on fatigue resistance [17–
20]. Four-point bending fatigue testing may be a suitable fatigue testing method to study the surface
integrity after machining, since the highest stresses from testing will be concentrated at the sample
surface.
This study investigates the effect of milling and EDM on high cycle fatigue of AISI 304 austenitic
stainless steel tube material. Fatigue resistance in relation to surface integrity is investigated using the
3
scanning electron microscopy (SEM) techniques, electron channelling contrast imaging (ECCI), electron
backscatter diffraction (EBSD), and energy dispersion spectroscopy (EDS), and hardness testing, and
RS measurements using X-ray diffraction (XRD).
2. Material and experimental procedure
The AISI 304 austenitic stainless steel used for this study was provided by Sandvik Materials Technology
in Sandviken, Sweden. The tube material was manufactured by hot extrusion cold pilgering and
solution annealed at 1060 °C for 15 minutes. Chemical composition of the steel in weight percentage
is C: 0.015, Si: 0.35, Mn: 1.2, Cr: 18.3, Ni: 10.3, W: 0.05, Cu: 0.3, Nb: 0.01, N: 0.07, and balance Fe. The
mechanical properties were i.e., yield strength, Rp0.2 = 210 MPa, tensile strength, Rm = 515 MPa,
Young´s modulus, E = 200 GPa.
Ten rectangular shaped samples were extracted from the tube using milling and EDM respectively.
Milling was performed with coolant and a Sandvik Model 390 cutting tool with a ø of 16 mm and new
inserts, 0.8 mm nose radius, rotation speed of 1200 rpm and a feed rate of 200 mm/min. A schematic
sketch of the milling process is shown in Figure 1 (a). EDM was performed using a ø 0.25 mm brass
wire with a cutting speed of 7.6 mm/min, 50 V, and 4.0 A. Surface roughness, Ra, measurements were
conducted using a Mitutoyo Surftest SJ-201M. Measurements of the milled and the EDM surface had
a Ra value of 0.81 and 2.27 respectively.
2.1 Fatigue testing
Four-point bending fatigue testing was done in a servo-hydraulic MTS machine using a Instron 8800
control system with a frequency of 15 Hz and a load ratio of R = 0.1. The experimental setup is shown
in Figure 1 (b). To fulfil the recommendations by Zhai et al. [21], samples measuring 10×10×80 mm, an
inner span of 13 mm, and an outer span of 58 mm for the test setup were used. A drop in displacement
of 3.5 % was used as the failure criterion. A test was considered a run-out (RO) if exceeding two million
cycles.
4
Figure 1: Schematic sketch of the (a) milling process, and (b) four-point bending setup.
2.2 Hardness testing and X-ray diffraction
A Struers DuraScan G5, following ISO 6507 and ASTM E384, equipped with a Vickers diamond was used
for hardness measurements with an applied load of 0.01 g. For hardness depth profiles 65 and 30
indentations were used for the milled and EDM samples respectively. Bulk hardness was measured
using 35 indentations.
X-ray measurements were performed using a four-circle goniometer Seifert X-ray machine, equipped
with a linear position sensitive detector and a Cr-tube. RS evaluations were conducted using the sin²ψ-
method [22] with the ɣ-Fe {220} diffraction peak, at 2θ ≈ 128.8°. X-ray elastic constants s1 and ½s2
were taken from reference [23]. The RS measurements conducted, used nine equally spread sin²ψ
values with ψ-angles between ± 55°. Material removal was done using a perchloric acid-base
electrolyte in a Struers LetcroPol-5 machine. No corrections were done for material removed.
2.3 Microscopy
Samples were prepared by grinding and mechanical polishing using a Struers Tegramin machine with
the parameter settings in Table 1. Samples were cleaned using water, soap and cotton after each
grinding operation. After each polishing cloth they were cleaned with water, soap and cotton followed
by ultrasonic cleaning in ethanol and a final cleaning with water, soap and cotton. A Hitachi SU-70 field
emission gun SEM, operating at 10 – 20 kV, was used to study the sample surface integrity and
deformation caused by the different cutting procedures. The analytical SEM-techniques used were
ECCI, EBSD, and EDS.
5
Table 1: Sample preparation steps.
Grinding paper or
polishing cloth
Grit size [µm] Load (each
sample) [N]
Time
[min]
Number
of
papers
Material
removed
[µm]
SiC paper 500 30 40 4.00 2 400
SiC paper 1200 15 40 4.00 1 60
SiC paper 4000 5 40 4.00 3 35
Silk cloth Diamond
suspension 3
30 8.00 10
Woven wool cloth Diamond
suspension 1
15 10.00 5
Rayon-viscose fibres
cloth
Diamond
suspension 0.25
15 15.00
Neoprene foam cloth Colloidal silica
suspension 0.04
15 5.00
Neoprene foam cloth Water 15 1.00
ECCI investigations were performed using a solid state 4-quadrant backscatter electron detector, an
acceleration voltage of 10-20 kV, and a working distance of 7-8 mm. Changes in the crystallographic
structure gives the speckled pattern observed with ECCI [24–26]. The probability of detecting back
scattering electrons varies with the rotational changes around any axis of the crystal. Local mis-
orientation, defects, and strain fields are shown as contrast variations because ECCI uses the
interaction between backscattered electrons and the crystal planes to generate contrast. This makes
ECCI is a good technique in order to investigate deformed materials [24–26]. It is not possible to
separate or quantify the contributions from elastic and plastic strain when simultaneously present.
6
Crystallographic orientation quantification was done using an OXFORD electron backscatter diffraction
(EBSD) detector. Sample configuration used: tilted 70°, a working distance of 20 mm, an acceleration
voltage of 15 kV and a step size of 0.5 µm. To evaluate the EBSD measurements the HKL software
Channel 5 was used. A mis-orientation (orientation difference between two neighbouring
measurement points) between 1.5o – 10o defines a LAGB while > 10o is regarded as a high angle grain
boundary. In the EBSD maps LAGB’s are represented with black lines, angles between 10o – 50o are
with red lines, and angles > 50° with blue lines. Non indexed points (zero solutions) are represented as
white dots.
Qualitative chemical composition measurement was done using an OXFORD EDS detector at a working
distance of 15 mm and an acceleration voltage of 20 kV.
3. Results
All milled samples were run-outs (RO) when tested with a maximum stress of 400, 410, 420, and 425
MPa. The two samples run with a maximum stress of 430 MPa failed after 346500 and 776000 cycles
respectively. The EDM samples were RO for all tested loads. Fatigue testing results are listed in Table
2.
Table 2: Fatigue testing results for milled and EDM surfaces.
Max stress
[MPa]
350 375 400 410 420 425 430 440 450 460 470
Milled - - RO RO RO RO, RO X, X - - - -
EDM RO RO RO RO RO - RO RO RO RO RO
Position tracking at minimum load revealed a significant difference in material behaviour for the two
surface conditions as shown in Figure 2. The starting value (zero) originates from a starting load of 0.6
kN, then the averaged position at minimum load over 1000 cycles are plotted for each data point.
7
Figure 2: Change in position from 0.6kN starting load. Position tracking at minimum load for milled
and EDM samples at the same loads.
3.1 Hardness testing and residual stress
Milling clearly increases the surface hardness as illustrated in Figure 3. Bulk hardness was HV0.01 185
±15. No changes in hardness could be seen for the EDM sample. The HV0.01 hardness for the EDM
material was 196 ±7 which is within the range of the bulk hardness.
Figure 3: Micro Vickers indentations for the milled sample illustrated with blue dots, the EDM
sample with orange dots, the un-affected bulk as a solid red line, and standard deviation with
dashed lines.
8
Figure 4 shows the biaxial stresses (solid lines) and FWHM-distribution (dashed lines) measured in
three different orientations for both the milled (a) and EDM (b) sample. Post milling, the surface RS
are ~ 500 MPa in tension and converges towards zero at a depth of approximately 30 – 40 µm. For the
milled material FWHM starts at 1.2° at the surface and drops to 0.6° at a depth of 30 µm, after which
the FWHM converges towards the bulk value of 0.35°. The EDM sample showed tensile RS of 200 MPa
at the surface which converged close to zero at a depth of ~ 10 µm. After which, the mean RS value
fluctuates around zero. FWHM at the surface was 0.75° and converged towards the un-affected bulk
value.
Figure 4: Residual stress and FWHM profile for the (a) milled and (b) EDM sample.
3.2 Microscopy
Three distinct zones can often be identified in Figure 5 (b) after milling: the severe plastically deformed
(SPD) zone, the heavily deformed zone, and the affected zone. Nano sized recrystallized grains
measuring approximately 40×90 – 75×150 nm in the SPD zone are shown in Figure 6 (a). The nano sized
9
grains had the longest side parallel to the cutting direction. The heavily deformed zone consists of a
mixture of deformation mechanisms in multiple directions and partially evolved nano sized grains as
seen in Figure 6 (c). Bending of mechanical twins (MT) is exemplified in Figure 6 (d). The affected zone
consists of everything from none to several deformation bands and a speckled pattern, see Figure 6
(b). There is still a rotation of the grains closest to the SPD. A speckled pattern, which is due to
variations in elastic and plastic strains, can always be found closest to the heavily deformed zone as
well as the extent of the MT.
Figure 5: The general microstructure of: (a) unaffected bulk material, (b) the three deformation zones
found after milling i.e., affected-, heavily deformed-, and the severe plastic deformation zone, and (c)
the typical deformation zones found in the samples after EDM. (d) – (f) Orientation imaging maps of
un-affected bulk, milled, and EDM material respectively. (g) IPF colour scheme. (h) LAGB-plot for the
un-affected bulk, milled, and EDM material. (i) – (k) Pole figures of un-affected bulk, milled, and EDM
material respectively.
10
Figure 6: Microstructure images all with the milled surface in the left-hand side showing: (a) The
severe plastic deformation zone after milling with recrystallized nano sized grains (some highlighted
in white). (b) The affected zone after milling. Speckled pattern and just a few deformation bands
located ~ 20 µm from the milled surface. Heavy deformation zone with a (c) mixture of partially
evolved nano sized grains and many active deformation systems and (d) highly dense deformation
twins.
EDM cuts through the material by partially melting and “flicking” away a fraction of the work piece,
the deformation seen is an effect of a melting process. The main microstructural differences between
milled and EDM material can be seen when comparing Figure 5 (b) and (c).
The effects of EDM on the surface integrity can be divided into the three different zones: EDM debris,
the re cast zone, and the discharge affected zone (DAZ), as depicted in Figure 5 (c). The surface consists
of an oxide scale, which varies in thickness, followed by the re cast zone where small black “dots” are
11
present. EDS mapping shows an enrichment of copper and zinc in the re cast zone, presented in Figure
7. This makes it possible to conclude that these “dots” are leftovers from the brass wire. DAZ consists
of the bulk material, not melted, but affected, by the heat from EDM, in where a pattern similar to a
dislocation cell occurs.
Figure 7: EDS mapping showing concentration of zinc and copper, leftovers from the brass wire, in the
regions rich on black dots.
LAGB results are shown in Figure 5 (h). Milling results in a small area with zero solutions in the SPD
zone due to nano sized grains and possibly edge effects. This results in a LAGB density of 1. The LAGB
density decreases close to the bulk value during the first 15 – 20 µm from the milled surface. Below
this depth, the LAGB density then continues to decrease reaching the same value as the non-deformed
material at a depth of 40 µm.
The LAGB density of the EDM material starts from 0.7 which indicates that the process does not induce
a severely plastically deformed surface with the associated strain hardened layer. The re cast material
from the work piece and wire causes a steep thermal gradient during cutting, resulting in the observed
LAGB density. EDM affects the material to a depth of ~ 15 µm, with most of the effect at the first five
microns from the surface.
4. Discussion
The existing research on the effects of EDM on fatigue life has mainly focused on tool steels [27,28],
superalloys [29], and Ti alloys [30,31].
12
The fatigue results obtained for the milled surface were as predicted, the material failed at a certain
load level. The EDM results, on the other hand, were more unexpected, they were all run-outs.
A fatigue limit does exist for the milled condition even though it was not fully established. This was not
the case for the EDM samples, instead they continued plasticising, and none of the samples failed prior
the two million cycle failure criterion over the whole stress range. There should be a difference in
fatigue strength between the milled and EDM condition. Since surface finish has been found to
dramatically change the fatigue strength as reported in [32], where polishing of AISI 316LN was found
to be detrimental to the fatigue properties compared to shot peening. The fatigue strength for the
milled condition was determined to be approximately 194 MPa. There is very little information
regarding four-point bending fatigue of stainless steels compared to rotary bending and uniaxial
fatigue testing. A summary of existing fatigue strength data at two million cycles for AISI 304, AISI 316,
and their derivatives is listed in Table 3 compared for the fatigue test stress amplitude for samples with
similar surface integrities. The obtained four-point bending results show greater strength for both the
milled and EDM conditions, compared to that found in literature [7]. Due to large scatter, no
conclusions can be drawn from three-point bending fatigue data [7–9]. The pure bending results
reported for the stronger alloy AISI 316 are significantly lower than ours [32]. Compared to rotary
bending [6,10,14,33] our results are roughly 95 MPa lower, which is not so surprising due to the testing
method. The results for the milled sample are in line with that of the uniaxial fatigue data for both 304
and 316 alloys [5,34–37].
13
Table 3: Fatigue strength data at two million cycles compared at maximum load for samples with
similar surface integrities. Numbers inside brackets () refer to tested material. The results cannot be
directly compared due to mean stress effects.
References
4-point bending
Uniaxial Rotary bending
3-point bending 4-point bending
Pure bending
σamp
[MPa] Milled
194 180 (304L) [34] 280 (304) [35] 185 (304L) [36] 370 (316) [37] 240 (316) [5]
233 (304) [10] 262 (316) [10] 290 (304) [33] 290 (304) [14] 366 (316L) [6]
100 (304) [9] 315 (304) [8] 172 (316L) [7]
115 (316L) [7] 220 (316) [32]
EDM > 212 RO
175 (316L) [7]
Three and four-point bend fatigue testing were conducted with electro-polished and ground notched
316L samples [7]. AISI 304 being the low-grade alloy when compared to AISI 316 still show similar
fatigue strength when comparing our milled condition specimens with the ground specimens in [7].
This is not the case when comparing our EDM condition specimens with the electro-polished
specimens in [7], even though the surface integrity of the two are fairly similar, the EDM samples never
fractured (all were run-outs). The fatigue results cannot be directly compared due to mean stress
effects.
Surface integrity is mainly tested during four-point bending fatigue. This is why a milled sample can be
treated as if the entire sample has been plastically deformed due to milling, EDM samples can instead
be treated as un-affected bulk material compared to uniaxial testing. The accumulation rate of inelastic
deformation in the EDM sample was likely too low in order to reach the strain limit needed to induce
a surface crack and final fracture. For the milled samples the accumulation rate of inelastic
deformation was sufficient enough to produce the dislocation density needed to induce a surface
crack.
The displacement curves in Figure 2 for the EDM conditions show that the curves decline over the
whole testing cycle which is likely due to ratcheting, no decline can be detected for the milled samples.
A change in material response was observed using position tracking, as shown in Figure 2. At one
14
hundred thousand cycles the milled samples reach a kind of “steady state” remaining to the end
compared to the EDM samples which continuously declined. Hysteresis loops for the two conditions
at same load level in Figure 8 revealed that ratcheting effects are greater in the EDM sample than the
milled. Ratcheting in austenitic stainless steels is common during uniaxial cyclic loading when the
nominal stress level is ≠ zero [11–13,38]. The ratcheting effect makes fatigue strength evaluation quite
tricky and is often disregarded. Even though four-point bending fatigue is not the most suitable way to
test AISI 304 due to the amount of ratcheting seen in Figure 8, it is very good when the surface
condition/integrity effects are of interest.
Figure 8: Hysteresis loops for cycle number 30, 100, 1000, 10000, 100000 and 1000000 at 420 MPa
load for milled and EDM samples. With increasingly larger cycle number, the hysteresis loop has
shifted to the left.
Microscopy
The microscopy investigation was performed prior to the four-point bending fatigue testing. The main
microstructural results are presented in Figures 5 and 6. In Figure 5 the general microstructure of un-
affected bulk, milled, and EDM material is shown in (a) – (c) respectively. Orientation imaging maps of
un-affected bulk, milled, and EDM material is shown in (d) – (f) together with an IPF colour scheme in
(g) as well as pole figures of un-affected bulk, milled, and EDM material respectively in (i) – (k). A LAGB-
15
plot for the un-affected bulk, milled, and EDM material is presented in (h). Figure 6 gives a closer view
of the different deformation mechanisms present in the milled samples e.g., recrystallized nano sized
grains, deformation bands, and MT.
Coolant was used during milling, resulting in a thin oxide layer on the sample surface. The oxide layer
was approximately 1 µm thick. Followed by recrystallized nano sized grains in the SPD shown in Figure
6 (a) and (c). In references [8,39–41] it is shown that multiple slip band and nano sized twins in several
activated slip systems is caused by machining, forming a grid of high dislocation density walls, from
where the recrystallized nano sized grains evolve due to the cutting tool heat generation. In the SPD
the LAGB is ~1, due to few indexed points.
In the heavily deformed zone, multidirectional deformation mechanisms are present. Mechanical
twins, shear bands, and slip bands have all been pointed out in the literature to be the main active
deformation systems [8,39–44]. The bent MT in Figure 6 (d) are caused by the strong shearing of the
work piece induced during milling. MT induced in the stress field will bend as a result of the step strain
gradient at the shear band [18]. The heavily deformed area shows a high LAGB density, and high
indexation, in contrary to the SPD. LAGB are illustrated with black lines in Figure 5 (e), some are short
and some are continuous band. In order to establish if these bands are twins or slip bands a higher
resolution is needed than that of the SEM. TEM investigations [8,39–41] on MT in austenitic stainless
steels have shown that the twin bands are built up by twining lamellas 10–30 nm in width with and
similar spacing between them. This being the reason for why the EBSD mappings could not fully prove
the existence of twins. The in-depth ECCI investigations, show that the majority of deformations bands
observed were actually MT bands. Pole figures of the affected, and heavily deformed areas, displays a
clear shearing of the grains with an increasing shearing towards the surface. Mostly scattered twin
lamella bands and lattice distortion is seen in the affected zone. The lattice distortion is induced by the
surface shearing, which contributes to the speckled pattern seen with ECCI in Figure 6 (b). There is a
good agreement in the affected depth seen with LAGB density plot and the ECCI.
16
EDM debris on the EDM sample surface has a thickness ranging between 0–20 µm and contains large
voids. Spallation is the reason for the zero measurement value. Below the debris is what we refer to
as the “re cast” layer as presented in Figure 5 (c). The black “dots” in the re cast layer were found to
be enriched with copper and zinc originating from the brass wire used for the EDM process, as well as
re-melted material emanating from the work piece. Beneath the re cast layer is the DAZ, DAZ is an
effect of the temperature gradient occurring in the work piece during cutting. EDS mapping and line
scans were conducted, in order to identify the source of the speckled pattern shown in the DAZ.
However, no changes in chemical composition in the DAZ could be seen. Therefore, the speckled
pattern seen using ECCI is not due to chemical changes. The differences in LAGB density shown in
Figure 5 (h), suggests that the speckled pattern is due to crystallographic changes, most likely in form
of dislocation structures or density. This is supported by the pole figure in Figure 5 (k) where the grains
closest to the surface are illustrated as clusters and not as the single points seen in the un-affected
bulk pole figure, Figure 5 (i). The speckled pattern observed with ECCI are most likely dislocation cells
[8,39,42]. From the LAGB density, the thickness of the DAZ is approximately 15 µm which is the same
depth as seen with the ECCI. The DAZ, having a thickness of approximately 15 µm, gradually changes
into un-affected material.
Hardness and residual stress profiles
Micro Vickers indentations of the milled sample showed that the milled surface and sub-surface were
strain hardened. The effect of strain hardening is illustrated in Figure 4, as the bulk value of 185 ±15
HV0.01 increases to approximately 350 HV0.01 at the milled surface. This hardness increase is well in line
with values reported for 316L [7,45]. The EDM sample did not show any evident hardening effects with
the value of 196 ±7 which is similar to the bulk hardness. Indentations must be placed far enough away
from the edge to avoid edge effects. Therefore, the values obtained closest to the edge should be used
with caution. The milled samples were expected to give an increase in fatigue strength due to strain
hardening compared to the EDM samples. Unsuspectingly, it was the opposite, the EDM samples were
17
all RO for our stress range and the milled samples showed a fatigue strength approximately at 194 MPa
at our failure criterion as presented in Table 2.
Tensile RS in the surface region are considered detrimental to the fatigue life. Austenitic stainless steels
are difficult to machine without generating tensile RS [46]. RS in the milled sample shows a drop from
~+500 MPa to approximately zero stresses at 30 µm. After which the stresses remain close to zero for
the whole measured depth. Similar values have been reported in [47,48]. In parallel with the change
in RS, we see a similar drop in FWHM showing that the dislocation density decreases, resulting in a less
strained material. The drop in both RS and FWHM for the milled sample is presented in Figure 3 (a).
Less can be said about in change in RS in the EDM sample due to large error bars, seen in Figure 3 (b).
Similar issues for RS measurements using X-ray diffraction on AISI 304 has been reported in [49]. Due
to the re cast layer both the RS and the FWHM values at the surface should not be treated as absolute.
The high tensile RS in the milled sample are in the vicinity of the tensile strength for AISI 304. The high
tensile RS and the highly strained surface layer in the milled sample are the reason for why a fatigue
strength can be observed. It is problematic to establish a fatigue strength using four-point bending for
the EDM sample due to lack of RS and a strained surface which result in a very ductile material
response.
This study shows that all techniques used revealed approximately the same affected depth for both
the milled and EDM sample. The techniques complement each other, resulting in a better
understanding of the materials state post milling and prior the fatigue testing.
5. Conclusions
Two different surface conditions, were obtained by milling and EDM, of AISI 304 are fatigue tested and
characterised. Surface integrity quantification was done using SEM techniques such as ECCI, EBSD, and
EDS. Four-point bending fatigue was used to determine the fatigue performance of the machined
surfaces. The post milling deformation depth was approximately 30 – 40 µm, which is < 0.4 % of the
total sample height. As a consequence, this highly strained region changes the fatigue behaviour.
18
• The milled samples exhibit a traditional well established fatigue behaviour, whereas the EDM
samples did not fractured.
• Due to the ratcheting effect observed in AISI 304, determining the fatigue strength by using
four-point bending fatigue does not seem to be a suitable testing method for EDM condition.
• LAGB density can be used to quantitatively measure plastic deformation.
• Grain shearing and subsequent grain rotation when milling results in bent mechanical twins
beneath the surface.
• Mechanical twins and deformation bands form cell structures at the cutting surface, from
where nano sized grains evolve due to the heat of the operation.
6. Acknowledge
The authors would like to acknowledge the Swedish Government Strategic Research Area in Materials
Science on Functional Materials at Linköping University (Faculty Grant SFO‑Mat‑LiU#2009‑00971) for
financial support. A special thanks to Senior Prof. Sten Johansson for SEM assistance and Sandvik
Materials Technology for providing the material.
7. References
[1] A. Maurotto, D. Tsivoulas, M.G. Burke, Surface Integrity in Dry Milling of 304L Steel: A
Parametric Study, Procedia CIRP. 13 (2014) 156–162. doi:10.1016/j.procir.2014.04.027.
[2] V. Chomienne, C. Verdu, J. Rech, F. Valiorgue, Influence of Surface Integrity of 15-5PH on the
Fatigue Life, Procedia Eng. 66 (2013) 274–281. doi:10.1016/j.proeng.2013.12.082.
[3] S.G. Acharyya, A. Khandelwal, V. Kain, A. Kumar, I. Samajdar, Surface working of 304L stainless
steel: Impact on microstructure, electrochemical behavior and SCC resistance, Mater. Charact.
72 (2012) 68–76. doi:10.1016/j.matchar.2012.07.008.
[4] S. Ghosh, V. Kain, Microstructural changes in AISI 304L stainless steel due to surface
machining: Effect on its susceptibility to chloride stress corrosion cracking, J. Nucl. Mater. 403
19
(2010) 62–67. doi:10.1016/j.jnucmat.2010.05.028.
[5] M. Kamaya, M. Kawakubo, Mean stress effect on fatigue strength of stainless steel, Int. J.
Fatigue. 74 (2015) 20–29. doi:10.1016/j.ijfatigue.2014.12.006.
[6] L. Ceschini, G. Minak, Fatigue behaviour of low temperature carburised AISI 316L austenitic
stainless steel, Surf. Coatings Technol. 202 (2008) 1778–1784.
doi:10.1016/j.surfcoat.2007.07.066.
[7] A. Laamouri, H. Sidhom, C. Braham, Evaluation of residual stress relaxation and its effect on
fatigue strength of AISI 316L stainless steel ground surfaces: Experimental and numerical
approaches, Int. J. Fatigue. 48 (2013) 109–121. doi:10.1016/j.ijfatigue.2012.10.008.
[8] C. Ye, A. Telang, A.S. Gill, S. Suslov, Y. Idell, K. Zweiacker, J.M.K. Wiezorek, Z. Zhou, D. Qian,
S.R. Mannava, V.K. Vasudevan, Gradient nanostructure and residual stresses induced by
Ultrasonic Nano-crystal Surface Modification in 304 austenitic stainless steel for high strength
and high ductility, Mater. Sci. Eng. A. 613 (2014) 274–288. doi:10.1016/j.msea.2014.06.114.
[9] N. Ben Fredj, H. Sidhom, Effects of the cryogenic cooling on the fatigue strength of the AISI
304 stainless steel ground components, Cryogenics (Guildf). 46 (2006) 439–448.
doi:10.1016/j.cryogenics.2006.01.015.
[10] J. Muñoz-Cubillos, J.J. Coronado, S.A. Rodríguez, Deep rolling effect on fatigue behavior of
austenitic stainless steels, Int. J. Fatigue. 95 (2017) 120–131.
doi:10.1016/j.ijfatigue.2016.10.008.
[11] X. Yuan, S. Fu, D. Yu, X. Chen, Temperature-Dependent Bending Ratcheting Behavior of a
316LN Stainless Steel, Mater. Eng. Performance2. 25 (2016) 274–279. doi:10.1007/s11665-
015-1839-9.
[12] G. Kang, Y. Liu, Z. Li, Experimental study on ratchetting-fatigue interaction of SS304 stainless
steel in uniaxial cyclic stressing, Mater. Sci. Eng. A. 435 (2006) 396–404.
doi:10.1016/j.msea.2006.07.006.
[13] S. Date, H. Ishikawa, T. Otani, Effect of ratcheting deformation on fatigue and creep-fatigue
20
life of 316FR stainless steel, Nucl. Eng. Des. 238 (2008) 336–346.
doi:10.1016/j.nucengdes.2006.09.009.
[14] M. Nakajima, Y. Nakamura, K. Suzuki, Y. Bai, Y. Uematsu, Effect of solution treatment after
nitriding on fatigue properties in type 304 stainless steel, Int. J. Fatigue. 68 (2014) 103–110.
doi:10.1016/j.ijfatigue.2014.05.013.
[15] B. Ravi Kumar, S. Sharma, B.P. Kashyap, N. Prabhu, Ultrafine grained microstructure tailoring
in austenitic stainless steel for enhanced plasticity, Mater. Des. 68 (2015) 63–71.
doi:10.1016/j.matdes.2014.12.014.
[16] D.Y. Jang, T.R. Watkins, K.J. Kozaczek, C.R. Hubbard, O.B. Cavin, Surface residual stresses in
machined austenitic stainless steel, Wear. 194 (1996) 168–173. doi:10.1016/0043-
1648(95)06838-4.
[17] X. Wu, X. Pan, J.C. Mabon, M. Li, J.F. Stubbins, An EBSD investigation on flow localization and
microstructure evolution of 316L stainless steel for Gen IV reactor applications, J. Nucl. Mater.
371 (2007) 90–97. doi:10.1016/j.jnucmat.2007.05.028.
[18] Q. Xue, J.F. Bingert, B.L. Henrie, G.T. Gray, EBSD characterization of dynamic shear band
regions in pre-shocked and as-received 304 stainless steels, Mater. Sci. Eng. A. 473 (2008)
279–289. doi:10.1016/j.msea.2007.04.048.
[19] R. M’Saoubi, L. Ryde, Application of the EBSD technique for the characterisation of
deformation zones in metal cutting, Mater. Sci. Eng. A. 405 (2005) 339–349.
doi:10.1016/j.msea.2005.06.002.
[20] J. Nellessen, S. Sandlöbes, D. Raabe, Effects of strain amplitude, cycle number and orientation
on low cycle fatigue microstructures in austenitic stainless steel studied by electron
channelling contrast imaging, Acta Mater. 87 (2015) 86–99.
doi:10.1016/j.actamat.2014.12.024.
[21] T. Zhai, Y.G. Xu, J.W. Martin, A.J. Wilkinson, G.A.D. Briggs, Self-aligning four-point bend testing
rig and sample geometry effect in four-point bend fatigue, Int. J. Fatigue. 21 (1999) 889–894.
21
doi:10.1016/S0142-1123(99)00084-5.
[22] I.C. Noyan, J.B. Cohan, Residual stress - Measurement by Diffraction and Interpretation, 1987.
[23] B. Eigenmann, E. Macherauch, Röntgenographische Untersuchung von Spannungszuständen
in Werkstoffen. Teil III. Fortsetzung von Matwiss. und Werkstofftechn. Heft 3/1995, S. 148-
160 und Heft 4/1995, S. 199-216, Materwiss. Werksttech. 27 (1996) 426–437.
doi:10.1002/mawe.19960270907.
[24] R.J. Kamaladasa, Y.N. Picard, Basic Principles and Application of Electron Channeling in a
Scanning Electron Microscope for Dislocation Analysis, Microsc. Scince, Technol. Apl. Educ.
(2010) 1583–1590.
[25] S. Zaefferer, N.-N. Elhami, Theory and application of electron channelling contrast imaging
under controlled diffraction conditions, Acta Mater. 75 (2014) 20–50.
doi:10.1016/j.actamat.2014.04.018.
[26] I. Gutierrez-Urrutia, S. Zaefferer, D. Raabe, Electron channeling contrast imaging of twins and
dislocations in twinning-induced plasticity steels under controlled diffraction conditions in a
scanning electron microscope, Scr. Mater. 61 (2009) 737–740.
doi:10.1016/j.scriptamat.2009.06.018.
[27] O.A. Abu Zeid, On the effect of electrodischarge machining parameters on the fatigue life of
AISI D6 tool steel, J. Mater. Process. Technol. 68 (1997) 27–32. doi:10.1016/S0924-
0136(96)02523-X.
[28] F. Ghanem, C. Braham, M.E. Fitzpatrick, H. Sidhom, Effect of near-surface residual stress and
microstructure modification from machining on the fatigue endurance of a tool steel, J.
Mater. Eng. Perform. 11 (2002) 631–639. doi:10.1361/105994902770343629.
[29] D. Trauth, F. Klocke, D. Welling, M. Terhorst, P. Mattfeld, A. Klink, Investigation of the surface
integrity and fatigue strength of Inconel 718 after wire EDM and machine hammer peening,
Int. J. Mater. Form. 9 (2016) 635–651. doi:10.1007/s12289-015-1249-4.
[30] S.L. Soo, M.T. Antar, D.K. Aspinwall, C. Sage, M. Cuttell, R. Perez, A.J. Winn, The effect of wire
22
electrical discharge machining on the fatigue life of Ti-6Al-2Sn-4Zr-6Mo aerospace alloy,
Procedia CIRP. 6 (2013) 215–219. doi:10.1016/j.procir.2013.03.043.
[31] F. Klocke, D. Welling, J. Dieckmann, Comparison of grinding and Wire EDM concerning fatigue
strength and surface integrity of machined Ti6Al4V components, Procedia Eng. 19 (2011) 184–
189. doi:10.1016/j.proeng.2011.11.099.
[32] E.R. de los Rios, A. Walley, M.T. Milan, G. Hammersley, Fatigue crack initiation and
propagation on shot-peened surfaces in A316 stainless steel, Int. J. Fatigue. 17 (1995) 493–
499. doi:10.1016/0142-1123(95)00044-T.
[33] K. Suzuki, Y. Nakamura, M. Nakajima, Y. Bai, Y. Uematsu, Improvement of Fatigue Properties
by Solution Treatment in Nitrided Type 304 Stainless Steel, Procedia Mater. Sci. 3 (2014) 627–
633. doi:10.1016/j.mspro.2014.06.104.
[34] L. Vincent, J.-C. Le Roux, S. Taheri, On the high cycle fatigue behavior of a type 304L stainless
steel at room temperature, Int. J. Fatigue. 38 (2012) 84–91.
doi:10.1016/j.ijfatigue.2011.11.010.
[35] I. Nikitin, M. Besel, Effect of low-frequency on fatigue behaviour of austenitic steel AISI 304 at
room temperature and 25 °C, Int. J. Fatigue. 30 (2008) 2044–2049.
doi:10.1016/j.ijfatigue.2008.02.005.
[36] D. Paquet, J. Lanteigne, M. Bernard, C. Baillargeon, Characterizing the effect of residual
stresses on high cycle fatigue (HCF) with induction heating treated stainless steel specimens,
Int. J. Fatigue. 59 (2014) 90–101. doi:10.1016/j.ijfatigue.2013.09.011.
[37] X. Xie, D. Ning, J. Sun, Strain-controlled fatigue behavior of cold-drawn type 316 austenitic
stainless steel at room temperature, Mater. Charact. 120 (2016) 195–202.
doi:10.1016/j.matchar.2016.08.031.
[38] L. Taleb, G. Cailletaud, Cyclic accumulation of the inelastic strain in the 304L SS under stress
control at room temperature: Ratcheting or creep?, Int. J. Plast. 27 (2011) 1936–1958.
doi:10.1016/j.ijplas.2011.02.001.
23
[39] J.Z. Lu, K.Y. Luo, Y.K. Zhang, G.F. Sun, Y.Y. Gu, J.Z. Zhou, X.D. Ren, X.C. Zhang, L.F. Zhang, K.M.
Chen, C.Y. Cui, Y.F. Jiang, A.X. Feng, L. Zhang, Grain refinement mechanism of multiple laser
shock processing impacts on ANSI 304 stainless steel, Acta Mater. 58 (2010) 5354–5362.
doi:10.1016/j.actamat.2010.06.010.
[40] F. Han, G.Y. Lin, Q. Li, R.F. Long, D.S. Peng, Q. Zhou, Influence of Different Deformation on
Microstructure and Properties of 304 Austenitic Stainless Steel, Adv. Mater. Res. 500 (2012)
690–695. doi:10.4028/www.scientific.net/AMR.500.690.
[41] H.. Zhang, Z.. Hei, G. Liu, J. Lu, K. Lu, Formation of nanostructured surface layer on AISI 304
stainless steel by means of surface mechanical attrition treatment, Acta Mater. 51 (2003)
1871–1881. doi:10.1016/S1359-6454(02)00594-3.
[42] T. Mayama, K. Sasaki, M. Kuroda, Quantitative evaluations for strain amplitude dependent
organization of dislocation structures due to cyclic plasticity in austenitic stainless steel 316L,
Acta Mater. 56 (2008) 2735–2743. doi:10.1016/j.actamat.2008.02.005.
[43] E.H. Lee, T.S. Byun, J.D. Hunn, M.H. Yoo, K. Farrell, L.K. Mansur, On the Origin of Deformation
Microstructures in Austenitic Stainless Steel : Part I — Microstructures, Acta Mater. 49 (2001)
3277–3287.
[44] E.H. Lee, T.S. Byun, J.D. Hunn, M.H. Yoo, K. Farrell, L.K. Mansur, on the Origin of Deformation
Microstructures in Austenitic Stainless Steel : Part I — Microstructures, Acta Mater. 49 (2001)
3269–3276.
[45] A. Yamamoto, T. Yamada, S. Nakahigashi, L. Liu, M. Terasawa, H. Tsubakino, Effects of surface
grinding on hardness distribution and residual stress in low carbon austenitic stainless steel
316L, ISIJ Int. 44 (2004) 1780–1782. doi:10.2355/isijinternational.44.1780.
[46] R. M’Saoubi, J.C. Outeiro, B. Changeux, J.L. Lebrun, A. Morão Dias, Residual stress analysis in
orthogonal machining of standard and resulfurized AISI 316L steels, J. Mater. Process.
Technol. 96 (1999) 225–233. doi:10.1016/S0924-0136(99)00359-3.
[47] C. Wiesner, Residual stresses after orthogonal machining of AlSl 304: numerical calculation of
24
the thermal component and comparison with experimental results, Metall. Trans. A. 23 (1992)
989–996. doi:10.1007/BF02675573.
[48] Y. Matsumoto, F. Hashimoto, G. Lahoti, Surface integrity generated by precision hard turning,
CIRP Ann. - Manuf. Technol. 48 (1999) 59–62. doi:10.1016/S0007-8506(07)63131-X.
[49] I. Altenberger, B. Scholtes, U. Martin, H. Oettel, Cyclic deformation and near surface
microstructures of shot peened or deep rolled austenitic stainless steel AISI 304, Mater. Sci.
Eng. A. 264 (1999) 1–16. doi:10.1016/S0921-5093(98)01121-6.